pretty much, yeah. I wonder why Tech Conferences get all those tech savvy people in and then fail at something a simple as a compressor/limiter on the audio.
We actually run bs1770gain over the audio, so that the levels are somewhat reasonable. Unfortunately we don't have the time and/or manpower to do manual audio vetting and correcting on 600+ videos.
FOSDEM the video and audio on location run unattended? I get the manpower point, I was just wondering about venue audio. The conferences I attend normally had „the audio person“ in charge of the audio on location. We then grab his signal, feed it through a limiter and then stream. Maybe that‘s an inspiration :)
@MathiasSchreiber: We do our very best to make sure the audio is reasonably good during the event. However, we have to cover and record 24 rooms, and (this year) 672 events in total. That's a ginormous amount, and so we have to rely on untrained volunteers to do most of the work during the event. If you have ideas on how to improve our audio quality next year, suggestions are always welcome (but probably best to do that through some other medium, the comment section on youtube isn't ideal for that ;-)
OMG! All those examples were equal parts mind-melting and infuriating at the same time. And, yeah, I would like service providers to stop rejecting '+' symbols.
The reason they reject them is you can add an infinite number of + symbols and still have email sent to a single address. test@gmail.com and test+@gmail.com will be registered as different emails but all emails will go to test@gmail.com. Think of this as them stopping multiple accounts being created with a single address.
@@xtc_w Not all email services do this, I think it's feasible that someone has a + in their main address. And I can just register a new email address which forwards to my main inbox anyways.
I expected him to go a bit more in depth into RFC5321 instead of just providing some samples asking whether it's valid or not. The rule itself is simple, why isn't it used in SMTP as well as RFC5231?
i could not catch that question but it was something with special chars in the domain part. IMO, since he's talking about smtp-level representation, not user-facing, it must be invalid. you can have ascii chars only in domain names, even if it starts with "xn--".
I believe the presenter may have misinterpreted the question @11:38 ... regardless of "how" you do it, injecting a user-provided string into *any* command input *can* have security concerns. If you're passing it to a mail program, some combination of characters may end up doing something unexpected. If you're passing it directly into a SMTP discussion ... same thing; you could very well trick the mail server into doing something completely unexpected with a properly (or "intentionally improperly") formatted address string ;)
Whats with a source-routed address (multiple @). It's still valid but not used anymore (you use BGP routers nowdays and do not need to provide a routing in the mail address anymore).
Damn this is some of the worst audio I have heard on YT or at a conference. If it sounded like this in person I feel sorry for anyone attending. I would have to leave a talk that sounded like this :-(
The stavros@io example is debatable. Although it's a syntactically valid email address, ICANN produced a report stating: "…the SSAC also recommends that the use of DNS resource records such as A, AAAA, and MX in the apex of a TopLevel Domain (TLD) be contractually prohibited where appropriate and strongly discouraged in all cases." - see: www.icann.org/en/system/files/files/sac-053-en.pdf
Technically speaking these are two independent standards, though. ICANN can prohibit use of A records in TLDs for Internet use, but you could create your own TLD for internal use. It just won't be internet accessible. But still may very well potentially function internally (although a lot of apps questionably ignore using a TLD anyway). But still standards compliant.
Am I missing something, or is there a big elephant in the room..? "stavros@io" appears to me as quite correct as long as io is a hostname on a local network, doesn't have to be a TLD...
So you think you can get away with using a $10 microphone at a conference A journey down audio quality
I thought this video was on validating emails, not raping my ears
RIP audio
pretty much, yeah. I wonder why Tech Conferences get all those tech savvy people in and then fail at something a simple as a compressor/limiter on the audio.
We actually run bs1770gain over the audio, so that the levels are somewhat reasonable. Unfortunately we don't have the time and/or manpower to do manual audio vetting and correcting on 600+ videos.
FOSDEM the video and audio on location run unattended?
I get the manpower point, I was just wondering about venue audio.
The conferences I attend normally had „the audio person“ in charge of the audio on location.
We then grab his signal, feed it through a limiter and then stream.
Maybe that‘s an inspiration :)
Another sidenote:
You need HARDWARE before you get it into the software thing. The signal gets distorted before it's in the computer :)
@MathiasSchreiber:
We do our very best to make sure the audio is reasonably good during the event. However, we have to cover and record 24 rooms, and (this year) 672 events in total. That's a ginormous amount, and so we have to rely on untrained volunteers to do most of the work during the event.
If you have ideas on how to improve our audio quality next year, suggestions are always welcome (but probably best to do that through some other medium, the comment section on youtube isn't ideal for that ;-)
OMG! All those examples were equal parts mind-melting and infuriating at the same time. And, yeah, I would like service providers to stop rejecting '+' symbols.
The reason they reject them is you can add an infinite number of + symbols and still have email sent to a single address. test@gmail.com and test+@gmail.com will be registered as different emails but all emails will go to test@gmail.com.
Think of this as them stopping multiple accounts being created with a single address.
@@xtc_w Not all email services do this, I think it's feasible that someone has a + in their main address. And I can just register a new email address which forwards to my main inbox anyways.
@@xtc_w it doesn't fix the problem because dots are optional for Gmail too.
I expected him to go a bit more in depth into RFC5321 instead of just providing some samples asking whether it's valid or not. The rule itself is simple, why isn't it used in SMTP as well as RFC5231?
i could not catch that question but it was something with special chars in the domain part. IMO, since he's talking about smtp-level representation, not user-facing, it must be invalid. you can have ascii chars only in domain names, even if it starts with "xn--".
To bad that the audio sucks.
I believe the presenter may have misinterpreted the question @11:38 ... regardless of "how" you do it, injecting a user-provided string into *any* command input *can* have security concerns. If you're passing it to a mail program, some combination of characters may end up doing something unexpected. If you're passing it directly into a SMTP discussion ... same thing; you could very well trick the mail server into doing something completely unexpected with a properly (or "intentionally improperly") formatted address string ;)
Please add subtitles to this.
I'm be like "that can't be it. where's the rest of it?"
Whats with a source-routed address (multiple @). It's still valid but not used anymore (you use BGP routers nowdays and do not need to provide a routing in the mail address anymore).
Sure in practice no one or at least virtually no one does it, but that doesn't mean it's not valid
Deep fried presentation
Damn this is some of the worst audio I have heard on YT or at a conference. If it sounded like this in person I feel sorry for anyone attending. I would have to leave a talk that sounded like this :-(
The stavros@io example is debatable. Although it's a syntactically valid email address, ICANN produced a report stating: "…the SSAC also recommends that the use of DNS resource records such as A, AAAA, and MX in the apex of a TopLevel Domain (TLD) be contractually prohibited where appropriate and strongly discouraged in all cases." - see: www.icann.org/en/system/files/files/sac-053-en.pdf
Technically speaking these are two independent standards, though. ICANN can prohibit use of A records in TLDs for Internet use, but you could create your own TLD for internal use. It just won't be internet accessible. But still may very well potentially function internally (although a lot of apps questionably ignore using a TLD anyway). But still standards compliant.
Am I missing something, or is there a big elephant in the room..? "stavros@io" appears to me as quite correct as long as io is a hostname on a local network, doesn't have to be a TLD...